Hey Product Hunt 👋 I'm super excited to share Markup with you—an open-source, AI-powered annotation tool for effortlessly transforming unstructured text into structured data for ML and NLP tasks like named-entity recognition. Markup learns as you annotate to predict and suggest complex annotations. In addition, Markup integrates seamlessly with widely-used and custom ontologies (e.g., UMLS) to help you quickly link concepts and ideas. Happy annotating 🚀

Markup has helped our clinical staff at Swansea University Health board annotate clinic letters at multiple sites. We have used it to produce gold standard annotations on our clinic letters across epilepsy, multiple sclerosis, cardiovascular and plastic surgery departments. The increase in AI tools is exciting, but for settings such as healthcare, we need a way to validate these tools against our own data. Markup allows this by making it easy for us to annotate our own data manually, and allow downstream models to be compared to those annotations. The best part about Markup is we can run it locally behind our own secure firewall as the code is open source! The GPT-4 annotation features looks great, and we are looking forward to making use of open-source alternative LLMs so that we can use them locally. Please see the following peer-reviewed published papers where we have used Markup in our analyses: https://www.frontiersin.org/arti... https://pubmed.ncbi.nlm.nih.gov/...

Markup is a great and easy to use product which has made a big difference to the annotation parts of our natural language pipelines. Highly recommended for text annotation

As an academic plastic surgery registrar involved in developing and validating a NLP pipeline for skin cancer, I am thrilled to share my experience with Markup. Markup has played a crucial role in effortlessly transforming unstructured text into structured data, specifically for named-entity recognition in our skin cancer research. One of the standout features of Markup is its ability to learn and adapt as you annotate. This intelligent system predicts and suggests complex annotations, making the annotation process smoother and more efficient. This functionality has been immensely valuable in our research, as it has significantly reduced the time and effort required for manual annotation. Furthermore, Markup seamlessly integrates with widely-used and custom ontologies, such as UMLS, enabling quick and accurate linking of concepts and ideas. This feature has been particularly useful in our skin cancer research, where the precise identification and classification of medical terms and entities are critical. Peer-reviewed published papers where we have utilised Markup: https://www.ncbi.nlm.nih.gov/pmc... https://academic.oup.com/bjs/adv...